Public Opinion and Climate Action: Insights from U.S. Survey Data (2010-2024)
Introduction
Climate change represents a gradual yet urgent global crisis. Despite widespread scientific consensus, public engagement with climate policy remains uneven across the United States. For policymakers and advocates, this variation presents both a challenge and an opportunity. How can outreach strategies be tailored to mobilize those who are concerned but inactive, and to increase support in communities where skepticism or disengagement is prevalent?
This project investigates how emotional concern (e.g., worry about global warming), perceived civic responsibility (e.g., belief that citizens should do more), and urgency perception (e.g., belief that climate impacts are imminent) relate to public support for climate policies. Rather than serving as a mere techinical exercise, this work addresses a substantive policy question on identifying the psychological and social drivers that may unlock higher levels of climate policy support. The aim is to provide evidence-based insights that can inform communication strategies, advocacy campaigns, and policy prioritization.
Dataset
Unit of Observation: Each row represents a U.S. state or county paired by year.
Number of Observations: 48,150 (51 states including the District of Columbia, and 3,156 counties over 2010-2024)
Number of Variables: 75 unique public opinion variables (e.g., beliefs, emotions, policy support), plus geographic identifiers and year
Key Variables
The following key variables are selected for analysis as they capture the psychological and social determinants that are widely cited in the literature on political activation, particularly in the context of climate behavior. Together, these variables provide a comprehensive and multidimensional view of the effective, social, temporal, institutional, and informational drivers of climate action support. Visualizing their geographic and temporal patterns offers valuable insights into where and why Americans may begin to take action on climate issues.
| Variable | Description |
|---|---|
year |
Specific year in which the public opinion estimates were generated, ranging from 2010 to 2024 |
GeoType |
U.S. state (including Washington D.C.) or county for each observation |
regulate |
Estimated percentage who support regulating carbon dioxide as a pollutant |
worried |
Estimated percentage who are worried about global warming |
timing |
Estimated percentage who think global warming will start to harm people in the U.S. now or within 10 years |
citizens |
Estimated percentage who think citizens should do more to address global warming |
discuss |
Estimated percentage who discuss global warming occassionally or often with friends and family |
mediaweekly |
Estimated percentage who hear about global warming in the media at least weekly |
The Survey Questions section of this document contains a complete list of additional variables and their corresponding survey questions used in the aggregation process.
Visualizations
Figure 1. Line Chart
This line chart visualizes the temporal trend (2010–2024) of public support for regulating carbon dioxide (CO2) as a pollutant across nine selected U.S. states. The vertical axis represents the percentage of survey respondents who support this policy, and the horizontal axis tracks the year.
The selection of states is based on their notable standing in 2024 in terms of both perceived civic responsibility and support for climate policy. These states represent consistently high (District of Columbia, Maryland, Massachusetts), moderate (New Mexico, Texas, Utah), or low (North Dakota, West Virginia, Wyoming) levels of engagement on climate issues from a civic and policy perspective.
For visual clarity, the line colors reflect each state’s worry level about climate change. Warmer fill tones indicate higher levels of concern. While the color mapping provides additional context, the groupings of states are based on civic and policy attitudes, not worry level alone.
The graph shows that states with high civic engagement and policy support exhibit strong and increasing support over time, while those in the low concern group have plateaued or declined in recent years. These divergent trends highlight geographic and ideological divides in public sentiment on climate policy.
Figure 2. Scatterplot
This scatterplot illustrates the relationship between climate concern and policy support across four U.S. Census Bureau regions: Northeast, Midwest, South, and West. Each point represents a state in a given year (2010-2024), with the x-axis showing the estimated percentage worried about global warming and the y-axis showing support for regulating CO2 as a pollutant.
The colored lines show smoothed trends (LOESS curves) for each region, with shaded areas representing confidence intervals around each trend line. Across all regions, greater concern correlates with higher support for regulation. The Northeast and South show the strongest association between climate worry and policy support. The Midwest displays moderate support that increases steadily with concern, and the West shows comparatively lower support for regulation at equivalent levels of concern, suggesting regional nuances in climate attitudes or policy preferences.
Figure 3. Animated Scatterplot
This animated scatterplot illustrates the relationship between climate concern and policy support across U.S. states from 2010 to 2024. Each point represents a state in a given year, with the x-axis showing the estimated percentage of residents worried about global warming and the y-axis showing support for regulating CO2 as a pollutant.
The color gradient of each point reflects the yearly slope coefficient from a linear model (regulate ~ worried), highlighting how strongly concern translates into policy support over time. Dark green indicates a weaker slope, gold represents moderate associations, and firebrick red denotes stronger relationships. A LOESS smoothing line is overlaid with a light confidence band, showing general trends over the years.
Figure 4. Heatmap
An interactive version of this visualization is available as a Shiny application here: https://cbuw46-0-0.shinyapps.io/climate-action/
This heatmap visualizes estimated support for climate-related beliefs and policies across all 50 U.S. states and the District of Columbia in the year 2024. Each row represents a state, and each column corresponds to a survey variable such as personal experience with global warming, levels of worry, and support for regulation or renewable energy funding. The color scale reflects the estimated percentage of people in each state who endorse a given belief or policy.
Model estimates are based on national survey responses on a wide range of questions on climate beliefs, perceived risks, policy support, and behaviors. The complete list of variables, along with the corresponding survey questions and responses used for their aggregation, is attached in the Survey Questions section of this document.
To enhance interpretability, both the state and variable axes were reordered based on their respective average response levels. States are arranged from those with the highest average support or concern (top) to those with the lowest (bottom). Variables on the x-axis are also sorted from widely supported beliefs and policies (right) to those with relatively lower average endorsement (left).
Brighter yellow tiles represent higher estimated support or agreement (closer to 80-85%), while darker purple and black tiles reflect lower agreement levels (around 30-40%).
The heatmap reveals several important patterns. Some states consistently show higher climate concern and policy support, while others cluster at the lower end. For instance, the District of Columbia and states like Massachusetts, Hawaii, Maryland, and New York show consistently high levels of climate concern and policy support. In contrast, many interior and southern states tend to show lower levels of concern and support.
Certain survey items (e.g., support for funding renewables or concern for future generations) show consistently high support, while others (e.g., personal experience with climate change or local disussion) show more variation or lower averages across states.
Figure 5. Choropleth Map with Bubble Plot
This choropleth map with an overlapping bubble plot visualizes two key climate-related indicators across U.S. states in 2024: how important global warming is to residents when voting, and how frequently they encounter climate information in the media.
Each state is shaded based on the estimated percentage of adults who say a candidate’s position on global warming is important to their vote. Brighter yellow and orange hues represent higher levels of perceived importance (up to 70%), while darker shades indicate lower levels (around 45%). Each state capital is marked with a bubble whose size and color reflect the estimated percentage of residents who are exposed to global warming information in the media at least once per week. Larger and brighter bubbles indicate higher levels of weekly media exposure.
Several regional patterns are observed. States in the Northeast and West Coast show both high media exposure and high climate importance in voting decisions. In contrast, many Midwest and Southern states show lower levels on both dimensions. These patterns suggest that access to climate information and political prioritization of climate issues are geographically associated.
Figure 6. Animated choropleth maps
This visualization presents a set of animated choropleth maps illustrating the percentage of residents in each state who report being worried about climate change (top) and the perceived urgency for national climate action (bottom) respectively, from 2010 to 2024. Green hues indicate lower levels of concern or sense of urgency, while warmer tones indicate higher levels.
Overall, the top map on state-level worry shows consistently higher percentages, reflected in the prevalence of yellow and orange tones compared to the bottom map. Both maps reveal a clear upward trend over time, with a marked shift from green to orange shades as highlighted by the animated year-to-year transitions.
Figure 7. Small-Multiple Maps
These small-multiple choropleth maps show county-level model estimates for two climate-related indicators across the United States from 2019 to 2024. The first set of maps at the top displays the estimated percentage of residents who think citizens should do more to address global warming. The second set of maps at the bottom shows the estimated percentage who discuss global warming occasionally or often with friends and family.
The color gradients reflect the percentage values for each county, with darker tones representing higher proportions and lighter tones indicating lower proportions. Across the top set of maps, there is a noticeable decline over time in the percentage of residents who think citizens should do more, reflected in the gradual lightening of map colors over time. In contrast, the bottom set shows no clear directional trend over the same period, with percentages remaining relatively stable and generally lower than the top set overall.
Next Steps
If additional time and resources were available, this project could be extended in several ways to deepen insights and strengthen policy relevance. First, the scope of variable could be expanded by incorporating behavioral indicators such as voting history, participation in climate-related events, or renewable energy adoption rates, allowing attitudes to be linked more directly with actual civic behaviors. Additional psychosocial predictors such as trust in institutions, political ideology, and personal experience with local climate impacts from other national or international surveys could also be integrated to enrich the analysis.
Second, the project could merge public opinion data with socioeconomic indicators (e.g., income, education levels) and environmental exposure measures (e.g., climate risk maps, disaster frequency) to identify structural factors influencing climate attitudes. This data integration would make it possible to conduct spatial analysis that accounts for similarities in public opinion across neighboring areas and potential spillover effects between them.
Third, longitudinal and predictive modeling could be applied to examine how changes in worry, urgency, and perceived peer norms precede shifts in climate policy support over time. Time-series cross-sectional analysis could uncover dynamic relationships while predictive models could forecast where increases in support are most likely under different outreach or policy scenarios.
Finally, targeted policy simulations could be conducted to estimate the potential impact of tailored communication campaigns in regions with low support but high latent concern. Such simulations could also assess the relative effectiveness of increasing urgency perception versus enhancing perceived civic responsibility in boosting climate policy support, offering practical guidance for designing evidence-based outreach strategies.
Conclusion
This project demonstrates how integrating geographic, temporal, and psychological dimensions of public opinion can illuminate the drivers of climate policy support in the U.S.. By leveraging high-quality, state- and county-level survey data from the Yale Program on Climate Change Communication, the analysis moves beyond simple descriptive statistics to highlight where and under what conditions public sentiment may be most responsive to targeted outreach.
The results indicate that emotional concern, civic responsibility, and perceived urgency each play distinct roles in shaping support for climate action, and that these effects vary across regions. The visualizations offer a clear and accessible way for policymakers, communicators, and advocates to identify high-potential areas for engagement.
Looking forward, extending this work with additional datasets, advanced modeling, and scenario-based simulations could help transform these insights into actionable strategies. By linking public opinion data to real-world outreach interventions, such research can contribute to more inclusive and effective climate policy-making. This approach addresses not only the urgency of the climate crisis, but also the democratic imperative of engaging the public in solutions.
Survey Questions
For mapping and analysis, similar response categories were combined. For example, “very worried” and “somewhat worried” were grouped as “worried.” Individuals who answered “don’t know” or left the question blank were not modeled separately.
| Variable | Description | Survey Question | Grouped Responses |
|---|---|---|---|
citizens |
Estimated percentage who think citizens should do more to address global warming | Do you think citizens themselves should be doing more or less to address global warming? | Much more; More |
congress |
Estimated percentage who think Congress should do more to address global warming | Do you think the U.S. Congress should be doing more or less to address global warming? | Much more; More |
consensus |
Estimated percentage who believe that most scientists think global warming is happening | Do you think most scientists think global warming is happening? | Yes |
discuss |
Estimated percentage who discuss global warming occasionally or often with friends and family | How often do you discuss global warming with your friends and family? | Often; Occasionally |
exp |
Estimated percentage who agree that they have personally experienced the effects of global warming | How much do you agree or disagree with the following statement: “I have personally experienced the effects of global warming.” | Strongly agree; Somewhat agree |
fundrenewables |
Estimated percentage who support funding research into renewable energy sources | How much do you support funding more research into renewable energy sources, such as solar and wind power? | Strongly support; Somewhat support |
futuregen |
Estimated percentage who think global warming will harm future generations a moderate amount or a great deal | How much do you think global warming will harm future generations of people? | A moderate amount; A great deal |
governor |
Estimated percentage who think their governor should do more to address global warming | Do you think your governor should be doing more or less to address global warming? | Much more; More |
gwvoteimp |
Estimated percentage who say a candidate’s views on global warming are important to their vote | How important will a candidate’s views on global warming be in determining your vote for President this year? Will it be the single most important issue, one of several important issues, or not important in determining your vote? | The single most important issue; One of several important issues |
happening |
Estimated percentage who think that global warming is happening | Do you think that global warming is happening? | Yes |
harmus |
Estimated percentage who think global warming will harm people in the US a moderate amount or a great deal | How much do you think global warming will harm people in the United States? | A moderate amount; A great deal |
human |
Estimated percentage who think that global warming is caused mostly by human activities | Assuming global warming is happening, do you think it is… ? | Caused mostly by human activities |
localofficials |
Estimated percentage who think their local officials should do more to address global warming | Do you think your local government officials should be doing more or less to address global warming? | Much more; More |
mediaweekly |
Estimated percentage who hear about global warming in the media at least weekly | How often do you hear about global warming in the media? | At least once a week |
personal |
Estimated percentage who think global warming will harm them personally a moderate amount or a great deal | How much do you think global warming will harm you personally? | A moderate amount; A great deal |
regulate |
Estimated percentage who support regulating carbon dioxide as a pollutant | How much do you support regulating carbon dioxide (the primary greenhouse gas) as a pollutant? | Strongly support; Somewhat support |
timing |
Estimated percentage who think global warming will start to harm people in the United now or within 10 years | When do you think global warming will start to harm people in the United States? | They are being harmed right now; In 10 years |
worried |
Estimated percentage who are worried about global warming | How worried are you about global warming? | Very worried; Somewhat worried |
Citation
Howe, Peter D., Matto Mildenberger, Jennifer R. Marlon, and Anthony Leiserowitz (2015). “Geographic variation in opinions on climate change at state and local scales in the USA.” Nature Climate Change, doi:10.1038/nclimate2583.
Marlon, J. R., Wang, X., Bergquist, P., Howe, P. Leiserowitz, A., Maibach, E., Mildenberger, M., and Rosenthal, S. (2022). “Change in US state-level public opinion about climate change: 2008–2020.” Environmental Research Letters 17(12), 124046
U.S. Census Bureau. (n.d.). “Census regions and divisions of the United States.” U.S. Department of Commerce, Economics and Statistics Administration, U.S. Census Bureau. Retrieved from https://www2.census.gov/geo/pdfs/maps-data/maps/reference/us_regdiv.pdf
Yale Program on Climate Change Communication. (2025). “Survey Questions.” Yale Climate Opinion Maps 2024. Retrieved from https://climatecommunication.yale.edu/visualizations-data/ycom-us/